Types of Learning

By: Anum Sidhu, MS Clinical Psychology

This article discusses main approaches to learning: classical conditioning, operant conditioning and observational learning. The two types of conditioning are a product of Behavioral Psychology, as it involves behavior observation and laboratory experiments. Conditioning is also referred as Stimulus-response learning (Hill, 2001). [1]

Classical Conditioning

Classical conditioning is also called Pavlovian conditioning as it was first discovered by a Russian physiologist Ivan P. Pavlov in 1900s. He was up to his research on digestion in dogs. His observation of salvation in dog laid the ground work on classical conditioning. Classical conditioning involves pairing of a neutral stimulus with a stimulus which already elicits a desired natural response. After learning, the conditioned (neutral) stimulus elicits a conditioned response without unconditioned stimulus (Mkzihe). [2]

Classical conditioning can be understood as learning about event series that occur independently of an individual’s actions in his environment. Especially, an individual learns that a previous event (also called stimulus) becomes an indication for a following event (Ploog, 2012). [3]. Phenomena of classical conditioning occurs when there is contingency between conditioned stimulus and unconditioned stimulus, thus results in change in magnitude and timing of conditioned response of human and animal subjects (Schmajuk, 2010). [4]

Turkkan (1989) integrated many research areas that have used classical conditioning as an explanatory model. The researcher alleged that data from a number of different disciplines show that classical conditioning processes has a larger role in the explanation of human and animal behavior than previously supposed. Data gathered from many seemingly unconnected phenomena like the placebo effects, relapse to drug or alcohol abuse by post addicts, and the immune system responses appear to include classical conditioning processes. [5]. While summarizing developments in the area of classical conditioning, Hout and Merckelbach (1991) said that previous studies have focused on following four misconceptions regarding what is classical conditioning:

  1. Firstly, classical conditioning does not result due to simple temporal pairing of conditioned stimuli with an unconditioned stimuli. Instead, a conditioned response occurs if a subject is able to forecast the incidence of a stimulus from the occurrence of another stimulus.
  2. Secondly, something which is learned through classical conditioning is not essentially a reaction to a cue, but it rather presents a probabilistic link between different stimuli.
  3. Thirdly, classical conditioning is not merely manifested in reactions mediated by autonomic nervous system (ANS), but also in many immunological factors, in evaluative judgments and in motoric behavior.
  4. Lastly, the nature of conditioned stimulus (CS) and unconditioned stimulus (US) is most of the time not a subject of indifference or irrelevance: particular combinations of conditioned stimulus and unconditioned stimulus produce more influential conditioning effects as compared to other combinations [6].

Procedural Learning

According to Poulos, Christian & Thompson (2008), classical conditioning is considered as an example of procedural learning. Procedural learning can be defined as a type of learning which involves acquiring a novel response or skill through experience. Procedural memory is a type of non-declarative memory and thus it does not depend on the attainment of knowledge which can be overtly recalled. Fear conditioning and Eye blink classical conditioning are the most passionately studied examples of classical conditioning in mammals. Moreover, classical conditioning provides an excellent model for understanding processes through which we generate adaptive behaviors. These behaviors thus reflect a change in our awareness of the fundamental structure of the world [7].

Operant Conditioning

Instrumental conditioning, also called operant conditioning, is a form of learning which allows an individual or an animal to adapt its actions in order to gain maximally from his environment while being rewarded for correct responses only (Touretzky and Saksida, 1997). [8]. The term operant conditioning was first coined by a behaviorist B.F. Skinner, that is why it is also referred as Skinnerian conditioning. Skinner believed in observable and external causes of behavior instead of internal motivations and thoughts (Cherry) [9].

Any event or object that increases the future probability of a behavior is called an operant reinforcer. On the other hand, any event that decreases the future probability of a behavior is called a punisher. Compared with classical conditioning, it includes a link between a response and consequences. B. L. Thorndike was the first to conduct experiment for determining relationship between a response and its consequences. Thus operant conditioning is based on Thorndike’s ‘Law of Effect’ which says that learning is determined by the effect which the response produces (Mkhize, 2008) [2]. Behaviors which are reinforced tend to be repeated in future (i.e. strengthened); behaviors which are not reinforced tend to die out in future or be extinguished (i.e. weakened). B. F. Skinner identified following types of responses, reactions or operants which can occur after a behavior.

  • Neutral operants: Responses from environment which can neither increase nor decrease the future probability of a behavior that is being repeated.
  • Reinforcers: Responses from environment that can raise the possibility of a behavior that is being repeated. Reinforcers can either be positive or negative. Positive reinforcement strengthens behaviors by giving a consequence which is rewarding for the individual. For example, a mother gives her child 1 chocolate each time he completes his homework (i.e. a reward), he will be more likely to repeat his behavior in future, thus behavior of completing homework is strengthened. Negative reinforcement refers to removal of an unpleasant reinforce. It strengthens behavior because removal of an aversive or adverse stimulus is ‘rewarding’ for the individual.
  • Punishers: Response from environment that decrease future probability of a behavior that is being repeated. Punishment always weakens behavior. Like reinforcement, punishment also involves application of an unpleasant stimulus. For example, a mother slapped her child for his misbehavior. Or removing a rewarding stimulus, for example, deducting his pocket money in order to punish his undesirable behavior. (McLeod, 2007) [10].
  • As punishment and rewards of behavior form the basis this learning, human brains gradually increase or decrease their reflexes for behaviors when they are rewarded or punished for doing those behaviors. Thus past experiences and consequences of an individual’s behavior help in anticipating future events. Human beings share this form of learning with many other animals, including invertebrates (Brembs, 2003) [11].

Comparison between neuronal responses in classical and instrumental conditioning

Tsitolovosky, Babkina and Shvedoc (2004) compared neuronal reactions during classical and instrumental conditioning under similar conditions. They said that it is necessary to distinguish between classical and instrumental conditioning, as in both cases the response to a paired or conditioned stimulus rises. They studied neuronal equivalents of instrumental and classical conditioning in the known neurons that were responsible for the self-protective closure of the pneumostome in the mollusk (Helix) under similar conditions.

During the classical conditioning part, a mollusk was punished after a physical (tactile) stimulus. However, during instrumental conditioning, a mollusk was punished when a known neuron did not produce a capacity for action as a result of a physical stimulus. Another physical stimulus, which was not ever paired to the unconditioned stimulus, was used as a discriminated stimulus. The researchers also compared the behavior of those known neurons during pseudo conditioning. They carried our experiments in a semi-intact planning. They examined how reactions to the physical and painful stimuli changed through different forms of training. It was found that dynamics of neuronal reactions to a conditioned physical stimulus were a lot more complex during instrumental conditioning as compared to classical conditioning and involved several phases.

They proved that response to a hurting stimulus during classical conditioning falls after a short-term initial rise. However, the neurons directing instrumental action remained highly sensitive to unconditioned stimulus during instrumental conditioning. Moreover, foreign neurons reduced their reactions to the unconditioned stimulus. The researchers tentatively concluded that instrumental and classical paradigms are primarily different at cellular level [12].

Observational Learning or Modeling

Observational learning or modeling is learning something through observation. It occurs as a function of observing, retaining and replicating new behaviors executed by others. For example, one can see kids learning a great deal from watching parents, peers and siblings. Even the behaviors they observe on television, video games and internet can also impact their thoughts, behaviors and actions. Observational learning is powerful; therefore, it is important to ensure that they observe the right kind of behaviors.

Observational learning was first identified in a famous and influential experiment of Albert Bandura, known as Bobo doll experiment. Bandura and his colleagues demonstrated one of the ways through which children learn aggression. Bandura’s social learning theory also suggests that learning occurs through observation and interaction with other people. The experiment involved 36 boys and 36 girls at Stanford University Nursery School. The children ranged in age between 3 and almost 6 years, and the average participant age was 4 years 4 months. Children were assigned to control, aggressive and non-aggressive group. The experiment involved exposing children to two different adult models; an aggressive model and a non-aggressive one. After witnessing the adult’s behavior, the children were then placed in a room without the model and were observed to see if they would imitate the behavior they had witnessed earlier. “Bandura made several predictions about what would occur. The results of the experiment supported three of the four original predictions.

  1. Children exposed to the violent model tended to imitate the exact behavior they had observed when the adult was no longer present.
  2. Bandura and his colleagues had also predicted that children in the non-aggressive group would behave less aggressively than those in the control group. The results indicated that while children of both genders in the non-aggressive group did exhibit less aggression than the control group, boys who had observed an opposite-sex model behavior non-aggressively were more likely than those in the control group to engage in violence.
  3. There were important gender differences when it came to whether a same-sex or opposite-sex model was observed. Boys who observed an adult male behaving violently were more influenced than those who had observed a female model behavior aggressively. Interestingly, the experimenters found in the same-sex aggressive groups, boys were more likely to imitate physical acts of violence while girls were more likely to imitate verbal aggression.
  4. The researchers were also correct in their prediction that boys would behave more aggressively than girls. Boys engaged in more than twice as many acts of aggression than the girls” (Cherry) [13].

Not all observed behaviors are effectively learned. Some factors play a role in whether learning is successful or not. Certain requirements and steps must be followed. The steps involved in the observational learning and modeling process are as follows:

  • Attention:
    In order to learn, one needs to pay attention. Anything that detracts one’s attention can have a negative effect on observational learning. If the model is interesting or there is a novel aspect to the situation, the individual is far more likely to dedicate his full attention to learning.
  • Retention:
    The ability to store information is also an important part of the learning process. Retention can be affected by a number of factors, but the ability to pull up information later and act on it is vital to observational learning.
  • Reproduction:
    Once the individual has paid attention to the model and retained the information, it is time to actually perform the behavior he has observed. Further practice of the learned behavior leads to improvement and skill advancement.
  • Motivation:
    Finally, in order for observational learning to be successful, one has to be motivated to imitate the behavior that has been modeled. Reinforcement and punishment also play an important role in motivation.

Heyes and Dawson (2007) conducted an interesting experiment on rats. They observed effects of observation on rats. Hungry rats were allowed to observe a conspecific demonstrator pushing a manipulandum, a joystick, to the right or to the left for food reward. Afterwards, they were allowed to access to the joystick. The effects of right-pushing vs. left-pushing observation experience were investigated on (1) response skill attainment, (2) reversal of left-right distinction, and (3) responding in the process of extinction. They found that rats that had observed left-pushing made more left responses during acquisition than rats that had observed right-pushing, and rats that had observed demonstrators pushing in the direction that had previously been reinforced took longer to reach criterion reversal and made more responses in extinction as compared to rats that had observed demonstrators pushing in the opposite direction to that previously reinforced. These results provided evidence that rats are capable of learning a response, or a response–reinforcer contingency, through conspecific observation [14].

According to Yi and Davis (2003), previous researches have indicated that modeling is an effective form of computer skills training. They developed and tested a different theoretical model of observational learning processes through which training programs based on influence computer skill performance. They alleged that observation based learning processes are signified as a second-order concept with these four dimensions: attention, motivation, production, and retention. They developed new measures for these dimensions. These measures had sound psychometric properties. They found out that observational learning processes significantly influence training outcomes. This new model provides a complete theoretical explanation of the mechanisms through which modeling-based interventions affect training outcomes [14].

Other Types of Learning

  • Perceptual learning – it involves the ability to learn to recognize those stimuli that have been seen before.
  • Motor learning – it involves formation of changes in motor system.
  • Relational learning – it involves connections between different areas of association cortex.
  • Spatial learning – it involves learning about the relations among different stimuli.
  • Episodic learning – it involves remembering sequences of events that one witnesses.

References:

  1. Hill, G. (2001). As level: Psychology through diagrams. Oxford university press. website address: http://books.google.com.pk/books?id=-rYkNmp4EPcC&pg=PA67&dq=classical+and+operant+conditioning&hl=en&sa=X&ei=YKlOUa-xPMvQ7AbA3oGACw&ved=0CDMQ6AEwAQ#v=onepage&q=classical%20and%20operant%20conditioning&f=false
  2. Mkzihe, N. (2008). Learning. In Nicholas, L (ed). Introduction to Psychology (2nd edition). UCT Press. Website address: http://books.google.com.pk/books?id=MP5X2SK2DCgC&pg=PA130&dq=classical+and+operant+conditioning&hl=en&sa=X&ei=YKlOUa-xPMvQ7AbA3oGACw&ved=0CDgQ6AEwAg#v=onepage&q=classical%20and%20operant%20conditioning&f=false
  3. Ploog, B. O. (2012). Classical Conditioning. Encyclopedia of Human Behavior (Second Edition). doi.org/10.1016/B978-0-12-375000-6.00090-2. Website address: http://www.sciencedirect.com/science/article/pii/B9780123750006000902
  4. Schmajuk, N. A. (2010). Mechanisms in classical conditioning: A computational approach. Cambridge University Press. doi: org/10.1017/CBO9780511711831.002. website address: http://ebooks.cambridge.org/chapter.jsf?bid=CBO9780511711831&cid=CBO9780511711831A009&tabName=Chapter
  5. Turkkan, J.S. (1989). Classical conditioning: The new hegemony. Behavioral and brain sciences, 12 (1). Website address: http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=6777200&fulltextType=RA&fileId=S0140525X00024572
  6. Hout, M.V.D., & Merckelbach, H. (1991). Classical Conditioning: Still Going Strong. Behavioural Psychotherapy, 19 (1).  doi.org/10.1017/S0141347300011514. Website address: http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=5850604&fulltextType=RA&fileId=S0141347300011514
  7. Poulos, A.M., Christian, K. M., & Thompson, R.F. (2008). Procedural Learning: Classical Conditioning. Learning and Memory: A Comprehensive Reference (3). doi.org/10.1016/B978-012370509-9.00121-2. Website address: http://www.sciencedirect.com/science/article/pii/B9780123705099001212
  8. Touretzky, D. S., & Saksida, L. M. ( 1997). Operant Conditioning in Skinnerbots. Adaptive behavior, 5 (3-4). doi: 10.1177/105971239700500302. Website address: http://adb.sagepub.com/content/5/3-4/219.short
  9. Cherry, K. Introduction to Operant Conditioning. About.com Guide. Website address: http://psychology.about.com/od/behavioralpsychology/a/introopcond.htm
  10. McLeod, S. A. (2007). B.F. Skinner | Operant Conditioning. Website address: http://www.simplypsychology.org/operant-conditioning.html
  11. Brembs, B. (2003). Operant conditioning in invertebrates. Current Opinion in Neurobiology, 13 (6). doi.org/10.1016/j.conb.2003.10.002. website address: http://www.sciencedirect.com/science/article/pii/S0959438803001570
  12. Tsitolovosky, L., Babkina, N., & Shvedoc, A. (2004). A comparison of neuronal reactions during classical and instrumental conditioning under similar conditions. Neurobiology of learning and memory, 81 (1). doi.org/10.1016/j.nlm.2003.08.002. Website address: http://www.sciencedirect.com/science/article/pii/S1074742703000923
  13. Cherry, K. Bobo Doll Experiment. About.com Guide. Website address: http://psychology.about.com/od/classicpsychologystudies/a/bobo-doll-experiment.htm
  14. Heyes, C. M. & Dawson, G. R. (2007). A demonstration of observational learning in rats using a bidirectional control. The Quarterly Journal of Experimental Psychology Section B: Comparative and Physiological Psychology, 42 (1). doi: 10.1080/14640749008401871. Website address: http://www.tandfonline.com/doi/abs/10.1080/14640749008401871
  15. Yi, M. Y., & Davis, F. D. (2003). Developing and Validating an Observational Learning Model of Computer Software Training and Skill Acquisition.Information Systems Research, 14 (2). doi:10.1287/isre.14.2.146.16016. Website address: http://infosys.highwire.org/content/14/2/146.short

For further reading:

  • Nicholas, L (ed). Introduction to Psychology (2nd edition). UCT Press. Website address: http://books.google.com.pk/books?id=MP5X2SK2DCgC&pg=PA130&dq=classical+and+operant+conditioning&hl=en&sa=X&ei=YKlOUa-xPMvQ7AbA3oGACw&ved=0CDgQ6AEwAg#v=onepage&q=classical%20and%20operant%20conditioning&f=false
  • Anderson, J. R. (2nd Edition). Learning and memory: An integrated approach. Website address: http://tocs.ulb.tu-darmstadt.de/89959310.pdf